Abstract
Circadian rhythms relate to multiple aspects of health and wellbeing, including physical activity patterns. Susceptible circadian regulation predisposes to circadian misalignment, poor sleep, sleep deprivation, increased sleepiness, and thereby sedentary behavior. Adolescents' circadian regulation is particularly vulnerable, and may lead to sedentary behavior. To investigate which factors associate strongest between physical activity (PA) and circadian behavior, we conducted multimodal circadian rhythm analyses. We investigate how individual characteristics of habitual circadian patterns associate with objectively measured PA. We studied 312 adolescents [70% females) (56% with delayed sleep phase (DSP)], mean age 16.9 years. Circadian period length, temperature mesor (estimated 24 h midline) and amplitude (difference between mesor and peak) were measured using distally attached thermologgers (ibutton 1922L, 3-day-measurement). We additionally utilized algorithm-formed clusters of circadian rhythmicity. Sleep duration, timing, DSP, and PA were measured using actigraphs (GeneActiv Original, 10-day-measurement). We found that continuous circadian period length was not associated with PA, but lower mesor and higher amplitude were consistently associated with higher levels of PA as indicated by mean Metabolic Equivalent (MET(mean)) and moderate-to-vigorous PA (MVPA), even when controlling for sleep duration. Separate circadian clusters formed by an algorithm also reflected distinct patterns of PA accordingly. Late sleepers and those with DSP were less likely to engage in MVPA compared to non-DSP and had more sedentary behavior. Adolescents who engage in higher levels or high-intensity PA have better circadian regulation, as measured by different objective methods including distal temperature measurements as well as actigraphy-measured sleep-wake behavior.